Problem solving and the scariest exam in the history of undergraduate education C.W.Wong, UCLA

• Why problem solving? • Why the old Mathematical exam? • Why the falling chain of Hopkins, Tait, Steele and Cayley? • Why does it conserve energy? • Conclusions

Kentucky, April 19, 2007 Why problem solving?

• Keh-fei is a good problem solver • Scientific research = problem solving • Can problem solving be taught? Optimistically: Yes •How? by the right people (teachers, mentors, peers, students) in the right places (institutions, facilities) at the right times (post Industrial Revolution) • Case study: The scariest (most competitive) exam in the history of undergraduate education = the old Cambridge Mathematical Tripos exam, 1780-1909 Why the old Cambridge Mathematical Tripos exam?

Model for written competitive exams in the English-speaking world

Cambridge University:

• 800th Anniversary in 2009 •81Nobel Prize winners worked or studied there, 1904-2005 • Isaac Newton studied and taught there: 1661-5, 1669-96 • Home of math. Physics in the 19th century: Clerk Maxwell studied and taught there: 1850-4, 1871-9 More about the Old Tripos (1):

•Tripos= three-legged stool on which the moderator sat in the original oral disputative exam (“wrangling”) • Old Tripos, 1780-1909: Required only for passing with honors Written exam Graduates ranked in Order of Merit Wranglers = students passing with first-class honors: (SW) = top student Second Wrangler (2W) = second best student Wooden Spoon (WS) = last student passing with honors • 1820-1909: The old Tripos became increasingly competitive. More about the Old Tripos (2):

• Lasting 2 weeks: 1st week: bookwork testing mastery of basic concepts, definitions, laws & proofs using Newton’s geometrical method 2nd week: problem papers often requiring new mathematical techniques applied to novel situations • Morning and afternoon exam papers, each usually 3 hours long: Each paper had 10-12 questions. Each question (15-18 minutes each) had several parts: theorem proof, simple application, difficult application. Usually all parts had to be answered correctly to score. • The 1881 Tripos: 18 three-hours papers with a total score of 33,541 marks: SW got 16,368 marks, 2W got 13,188, WS got 247. • Examiners believed that marks were proportionate to the abilities of the lower-placed candidates, but did not do full justice to the highest. Famous wranglers (1): Only five exceptional mathematicians:

• J.J. Sylvester (2W, 1837) •A.Cayley(SW, 1842): wrote 967 papers, 1841-95 (Euler wrote only 800 papers) •W.K. Clifford (2W, 1867): Cayley’s first student • G.H. Hardy (4W, 1898): collaborated with Ramanujan • J.E. Littlewood (SW, 1905): collaborated with Hardy Famous wranglers (2):

Many distinguished physicists thrived in the competitive atmosphere of Cambridge & the fellowship of the Royal Society:

•G. Stokes (SW, 1841): fluid dynamics, optics • W. Thomson (Lord Kelvin) (2W, 1845): trans-Atlantic cable, thermodynamics, Kelvin temperature scale • J. Clerk Maxwell (2W, 1854): kinetic theory, Maxwell equations (Perhaps greatest physicist after Newton & Einstein) • J.W. Strutt (Lord Rayleigh) (SW, 1865): discovered argon; gases • J.J. Thomson (2W, 1880): discovered electron; electricity in gases • William Bragg (3W, 1884): X-ray crystallography Coaches of the old Tripos:

Students prepared for the exam by taking private lessons from coaches.

• Coach = tutor/trainer, named after the stagecoach (Oxford undergraduate slang, 1830’s) • Coach = closed carriage, named after the Hungarian village Kocs where this carriage originated • Cambridge Students learned to solve problems against the clock from their coaches. (Training in memory and mental discipline.) More about the Old Tripos (3):

• Textbooks: were written directly or indirectly for students: Whewell (1823), Tait & Steele (1856), Routh (1898). •Collectionsof Tripos problems and solutions: Wright (1827-31), Wolstenholme (1867, 1878) and private collections of coaches • Tripos examiners: Always top Wranglers, usually young (within 4-8 years of graduation). Many were coaches. • The “arms race”, 1820-1909: Students were better prepared by coaches and by using textbooks and problem collections, but the problem papers became increasingly more difficult. 1828 Tripos: 4 days, 8 papers 1881 Tripos: 9 days, 18 papers • Tripos stress: Cakewalk: Cayley (SW, 1842), Rayleigh (SW, 1865) Had enough: Kelvin (2W, 1845) Nerves: Maxwell (2W, 1854), J.J. Thomson (2W, 1880) Nervous breakdown: James Wilson (SW, 1859) Could not walk 50 yards afterwards Took 3 months to recuperate & only by forgetting all Cambridge mathematics The modern math Tripos exams:

• 1909: Order of merit was abandoned because more students avoided the math Tripos by taking the easier Tripos exams in other subjects. • Present-day Math. Tripos: 3 years for the BA degree: with four 3-hour papers per year (called Parts IA, IB and II) on different mixed subjects Tripos Part III taken in the 4th year: 6 long (9 short) courses, each with a 3-hour (2-hour) exam Why the falling chain of Hopkins, Tait, Steele and Cayley?

Excessive respect for Newton’s methods retarded British mathematics and dynamics in the 18-19th centuries. It also caused mistakes in Tripos solutions: Example: Chain falls down link by link from a coil on a table. Incorrect energy-nonconserving (or inelastic) solution was given by coach William Hopkins, and published by

• Tait and Steele (SW & 2W, 1854), 1856: First textbook on problem

•Cayley(SW, 1842), 1857: First paper on problem • Wolstenholme (3W, 1850), 1878: Problem collection • Jeans (2W, 1880), 1907: Textbook •Lamb (4W, 1898), 1914: Textbook • Sommerfeld, Mechanik, 1943: Textbook • All the physics and engineering textbooks on mechanics that I have checked Mistakes in Tripos solutions:

Surprising after scrutiny by generations of teachers, coaches and students

• Hopkins: coached Stokes, Kelvin, Maxwellk, Tait, Todhunter • Cayley: Perhaps the best pure problem solver in 19th century Cambridge & the best mathematician of Victorian Britain • Sommerfeld: One of the best teachers of physics 4 Ph.D. students won Nobel Prizes: Pauli, Heisenberg, Debye, Bethe 2 postdocs won Nobel Prizes: Pauling, Rabi Correct energy-conserving (or elastic) solution has been given by

• C.W. Wong and K. Yasui, 2006: First correct theoretical treatment • C.W. Wong, S.H. Youn and K. Yasui, 2007: First indirect experimental confirmation Why does it conserve energy?

Mathematical picture:

ρ x 2 Lagrangian: L(x,v) = xv 2 + ρ g 2 2 d (ρ xv) = ρ xa + ρv 2 Lagrange equation: dt ∂L s = = −ρ gx + ρv 2 ∂x 2 Solution: g ⎧g/3 ⎧0 inelastic, a = = ⎨ , if s=⎨ 3−s ⎩g/2 ⎩1 elastic. Physical pictures Inelastic picture of Hopkins, Tait, Steele, Cayley and Sommerfeld: Falling link sticks to falling chain in a totally inelastic collision (Carnot’s theorem) Elastic picture of Wong and Yasui: Falling link gains energy when it breaks off from stationary chain segment. Falling link loses energy when it joins falling chain segment. The whole process is elastic. Conclusions:

•Best problem solvers have innate abilities, but are influenced by people, places and times. Case study: Famous problem solvers who took the Math Tripos exam in Cambridge (home of Newton and Maxwell)

in Maxwell’s time (right after the Industrial Revolution). • Original research is enhanced by fierce competition => The elite world of meritocracy, evolution, capitalism, politics. Lesson? To foster competition, only the best students should be graded A+. • Retention of learning is helped by rewards of pleasure => The feel-good world of democracy, socialism, Heaven (the Greatest Society) Lesson? To maximize pleasure & memory, every student should be graded A+. • Grading culture, begun in 19th century Cambridge, flourishes to this day. •The physics in this talk? There is a mistake in the accepted solutions of Tripos problems, after 150 years of careful scrutiny by generations of mathematicians and physicists.